NiCE-Metrigy Study: Trust, Not Tech, Drives AI Adoption in CX
New NiCE and Metrigy Research findings show consumers are more willing to engage with AI in customer service, but that openness depends on trust in the system's accuracy and transparency.
NiCE and Metrigy Research have released new findings showing that consumers are growing more open to artificial intelligence in customer service interactions, with trust emerging as the deciding factor in whether that openness translates into a better experience. The joint research frames AI adoption in customer experience not as a simple technology rollout, but as a trust-building exercise between brands and the people they serve.
The study points to a shift in consumer sentiment: rather than resisting AI-driven service outright, more customers appear willing to engage with it when they believe it will be accurate, transparent and genuinely helpful. That willingness, according to the research, is conditional — it rises or falls depending on how confident customers feel in the systems handling their requests.
For CX and digital transformation leaders, the finding reframes the AI conversation. The technical question of whether to deploy AI in service channels is increasingly settled; the harder question is how to earn and keep customer trust once it's live. That shifts investment priorities from pure automation capability toward transparency, reliability and the design choices that signal competence to the end user.
It also has implications for how organizations sequence AI rollouts. A trust-first approach suggests leaders should treat early AI interactions — chatbots, voice assistants, agent-assist tools — as trust-calibration moments rather than cost-cutting moves, since a poor first experience can suppress adoption even when the underlying technology works well.
Openness to AI is not the same as confidence in AI, and conflating the two is where many CX programs go wrong.
Most organizations measure AI success by deflection rates and cost savings, but the research points to a different scoreboard: perceived reliability. Trust in service AI isn't earned through capability alone — it's earned through consistent, legible behavior that customers can predict and verify. Operators should stop asking "can our AI handle this interaction?" and start asking "would a skeptical customer trust this answer enough to act on it?" That's a design and transparency problem, not a technology problem, and it's where the real competitive advantage will sit over the next few years.